business
Original Article
Real-World Corticosteroid Utilization Patterns in Patients with Metastatic Castration-Resistant Prostate Cancer in 2 Large US Administrative Claims Databases Marie-Hélène Lafeuille, MA; Jonathan Gravel, MSc; Amanda Grittner, MA; Patrick Lefebvre, MA; Lorie Ellis, PhD; R. Scott McKenzie, MD Background: Prostate cancer is the most common noncutaneous malignancy in men in the United States. Patients with metastatic castration-resistant prostate cancer (mCRPC) may be treated with secondary hormonal therapy or with chemotherapy, and potentially with concomitant corticosteroids. Corticosteroids can help manage the side effects of chemotherapy and secondary hormonal therapy and ameliorate prostate cancer–related symptoms, although corticosteroids are also associated with adverse effects. With an increasing number of available treatment options for mCRPC, evaluating the real-world concomitant use of corticosteroids in this patient population is important. Objective: To evaluate the utilization patterns of corticosteroids for the treatment of patients with mCRPC based on real-world data from 2 large claim databases. Methods: This retrospective analysis included medical and pharmacy claims from 2 large publicly available healthcare claims databases covering more than 31 million individuals to identify treatment patterns in adult patients with mCRPC. A total of 2593 patients with mCRPC were identified in data set 1 and 626 patients in data set 2 between 2005 and 2011. The appropriate treatment for castration-resistant prostate cancer (CRPC) was defined as chemotherapy, an antiandrogen, an adrenal androgen blocker, or estrogen. The index date was the date of the first CRPC treatment or the first metastasis diagnosis, whichever occurred later. The observation period spanned from the index date to the end of health insurance eligibility. Study end points included population characteristics, the distribution of mCRPC therapies, and corticosteroid utilization patterns. Results: The study population came from the 2 data sets and included 3219 men who were treated for mCRPC. Bone and lymph nodes were the predominant metastatic sites. Bicalutamide was the most common secondary hormonal therapy, and docetaxel was the most common chemotherapy used for these patients. Overall, 73.4% of the patients in data set 1 received concomitant corticosteroids, as did 71.6% of patients in population 2 during the entire period from the index date to the end of eligibility date. In addition, 62.8% and 60.4% of patients, respectively, received concomitant corticosteroids during the secondary hormonal therapy period, and 93.8% and 95.1% of patients, respectively, received concomitant corticosteroids during the chemotherapy period. Similar patterns of corticosteroid use were observed across geographic areas of the United States. Conclusion: This study shows consistently similar utilization patterns of corticosteroids in patients with mCRPC in 2 large national databases. Using real-world data to inform concomitant corticosteroid use in the treatment of patients with mCRPC may assist healthcare providers with treatment selection and with sequencing decision. Future research is warranted to investigate evolving treatment options for patients with mCRPC.
Stakeholder Perspective, page 315
Am Health Drug Benefits. 2013;6(6):307-316 www.AHDBonline.com Disclosures are at end of text
Ms Lafeuille is Senior Economist, Mr Gravel is Economist, Ms Grittner is Economist, and Mr Lefebvre is Vice President, Groupe d’Analyse, Ltée, Montreal, Quebec, Canada; Dr Ellis is Associate Director and Dr McKenzie is Senior Director, Janssen Scientific Affairs, Titusville, NJ.
Vol 6, No 6
l
July/August 2013
www.AHDBonline.com
l
American Health & Drug Benefits
l
307
business
Key Points Corticosteroids can help manage the side effects of chemotherapy or secondary hormonal therapy in patients with metastatic castration-resistant prostate cancer (mCRPC) and ameliorate prostate cancer–related symptoms. ➤ Real-world data on corticosteroid use and treatment patterns are infrequently reported in patients with mCRPC. ➤ Using real-world evidence from medical and pharmacy claims in 2 large claims databases covering >31 million lives, this study identified corticosteroid utilization patterns in patients with mCRPC. ➤ Results show that, in actual practice, the use of concomitant corticosteroid therapy in patients with mCRPC is consistent with the drug’s labeling information across US regions. ➤ In this study, patients with mCRPC were exposed to corticosteroid therapy approximately 33% of the time between 2005 and 2011. ➤ Such real-world data may offer providers positive feedback regarding their treatment selection and sequencing decisions for this patient population. ➤
P
rostate cancer is the most frequently diagnosed noncutaneous malignancy in males and is a leading cause of cancer-related morbidity and mortality among men in the United States.1,2 The American Cancer Society has estimated that approximately 238,600 new cases of prostate cancer will be diagnosed in 2013, and that approximately 29,700 men will die of the disease.3 Many patients will be cured with a primary therapy of surgery, radiation therapy, and potential adjuvant hormonal therapy, but a subset of patients will experience disease progression with metastases spreading to other organs or to tissues in the body, such as to the bladder or bone. For such patients, androgen-deprivation therapy (ADT) is a widely accepted treatment option. Disease progression despite ADT and castrate levels of testosterone is considered castration-resistant prostate cancer (CRPC).4,5 Up to 85% of patients with CRPC are estimated to have metastatic CRPC (mCRPC) and may be considered for secondary hormonal therapy or for chemotherapy.6,7 Historically, treatment for mCRPC was predominantly palliative; however, recent pharmacologic advances have led to increased treatment options showing survival benefits in addition to pain reduction and improvements in quality of life.7,8 An early trial showed that mono
308
l
American Health & Drug Benefits
l
therapy with prednisone may contribute to pain relief and a reduced need for analgesics in a subset of patients.9 In addition, several treatment regimens approved by the US Food and Drug Administration (FDA) for mCRPC involve corticosteroids. Prednisone in combination with mitoxantrone has been associated with pain reduction and with improvement in quality of life in men with advanced, hormone-refractory prostate cancer, but not with improved survival.10 Treatment with docetaxel in combination with prednisone was the first chemotherapy regimen shown to increase survival among patients with mCRPC.10 Men receiving treatment every 3 weeks with docetaxel plus a daily dose of prednisone had a 24% lower probability of death (hazard ratio, 0.76) compared with men treated with mitoxantrone plus prednisone.10 More recent FDA-approved treatment regimens administered with concomitant corticosteroids include abir aterone acetate in combination with prednisone, an oral androgen biosynthesis inhibitor for the treatment of patients with mCRPC, and cabazitaxel in combination with prednisone, a cytotoxic regimen for patients with metastatic hormone-refractory prostate cancer whose disease progresses and who have previously received docetaxel.11,12 Other recently approved treatment regimens without specific corticosteroid requirements include sipuleucel-T, an autologous cellular immunotherapy for asymptomatic or minimally symptomatic patients with mCRPC, and enzalutamide, an androgen receptor inhibitor for patients with progressed mCRPC who have previously received treatment with docetaxel.13,14 Corticosteroids can help manage the side effects of chemotherapy and secondary hormonal therapy and can ameliorate prostate cancer–related symptoms, although they are also associated with adverse effects. An increasing number of treatment options for mCRPC included in the National Comprehensive Cancer Network guidelines involve corticosteroids, but real-world data on corticosteroid use and treatment patterns have been reported infrequently in this patient population.15 The goal of the present study was to address this gap by evaluating the real-world corticosteroid treatment patterns for patients with mCRPC in the United States. A retrospective study design was used to describe the treatment patterns of patients with mCRPC.
Methods Data Source This retrospective study used data from the Truven Health Analytics MarketScan Research Databases (quarter 1 of 2005 through quarter 1 of 2011) and from the Optum Insight Clinformatics Data Mart database (quarter 1 of 2005 through quarter 2 of 2012). The Truven data-
www.AHDBonline.com
July/August 2013
l
Vol 6, No 6
Corticosteroid Utilization in Patients with mCRPC
Figure 1 Study Design
Start of data availability: Q1 of 2005
Start of eligibility
ADT therapy
Index date Initiation of first CRPC therapy or first diagnosis of metastases, whichever was later
Baseline period: 6 mos before index date
End of eligibility
End of data availability: Q1 of 2011a/ Q2 of 2012b
Observation period
Data availability for data set 1 (Truven Health Analytics MarketScan) ended in Q1 of 2011. Data availability for data set 2 (Optum Insight Clinformatics Data Mart) ended in Q2 of 2012. ADT indicates androgen-deprivation therapy; CRPC, castration-resistant prostate cancer; Q, quarter.
a
b
base (data set 1) includes data from 45 large employers, health plans, and government and public organizations covering more than 18 million lives through a commercial insurer or through Medicare. Enrollees in the database include employees, dependents, and retirees in the United States with primary or Medicare supplemental coverage through privately insured fee-for-service, point of service, or capitated health plans. All census regions are represented, although most enrollees are from the South and North Central (ie, Midwest) regions. The Optum database (data set 2) contains claims from more than 13 million privately insured individuals covered by 69 self-insured Fortune 500 companies with locations in all census areas of the United States. Data are available for all company beneficiaries, including employees, spouses, dependents, and retirees. Both databases contain information from integrated, individual-level claims of commercial health insurances, including medical claims for outpatient and inpatient services, pharmacy claims for prescription drugs, and eligibility information on demographics and health plan characteristics. The data were deidentified in compliance with the Health Insurance Portability and Accountability Act of 1996 to preserve patient confidential ity. Institutional Review Board approval and informed consent were not required for this study.
Study Design and Population Selection Adult patients who were treated for mCRPC after ADT, defined as either having at least 1 dispensing of luteinizing hormone-releasing hormone (LHRH) or an orchiectomy, who had at least 1 diagnosis of a malignant neoplasm of the prostate (International Classification of Diseases, Ninth Revision [ICD-9], 185.xx) and at least 1 additional diagnosis of any history of prostate cancer
Vol 6, No 6
l
July/August 2013
(ICD-9, 185.xx or V10.46) within 1 year of at least 1 metastatic diagnosis (ICD-9, 196-199) were included in the study population. Eligible treatments for CRPC included chemotherapy (eg, docetaxel, doxorubicin, estramustine phosphate, etoposide, mitoxantrone, cabazitaxel, paclitaxel, vin blastine, carboplatin, and vinorelbine); antiandrogens (eg, bicalutamide, flutamide, and nilutamide) unless initiated within 90 days of LHRH initiation; adrenal androgen blockers (eg, ketoconazole and aminoglutethimide); and estrogens (eg, diethylstilbestrol, conjugated estrogens, and other estrogens). To ensure that metastases stemmed from prostate cancer, men with another cancer diagnosis (ICD-9, 140.xx-184.xx, 186.xx-209.xx) within 1 year of the first metastatic diagnosis were excluded from the sample. Continuous eligibility for health insurance was required for at least 1 year before the first metastatic diagnosis and at least 6 months before the first CRPC treatment. To obtain a sufficiently long observation period, patients who had an index date less than 3 months before the end of the availability of the data (ie, cutoff date) were excluded from the analysis. Figure 1 illustrates the study design. The index date was defined as either the initiation of the first CRPC therapy or the date of the first diagnosis of metastases, whichever occurred later. The 6 months preceding the index date constituted the baseline period. The observation period spanned from the index date until the end of the individual patient’s health insurance eligibility or until the end of data availability.
Study End Points Population baseline characteristics The demographic and clinical characteristics of the
www.AHDBonline.com
l
American Health & Drug Benefits
l
309
business
study population were evaluated during the baseline period. These characteristics included age, payer type, region, year of index date, clinical history (including score on the Quan-Charlson comorbidity index [CCI]16,17), sites of metastases at the first metastatic diagnosis, and length of observation period.
Evaluation of mCRPC therapies The distribution of mCRPC therapies was evaluated to assess the treatment patterns of patients with mCRPC. The therapies were stratified into secondary hormonal therapies and chemotherapies. Secondary hormonal therapies included antiandrogens, excluding any antiandrogens initiated within 90 days of the initiation of LHRH therapy; adrenal androgen blockers; and estrogen. Relative frequencies for the first treatment of mCRPC and for all treatments during the observation period were calculated. Evaluation of corticosteroid use Because corticosteroid utilization may be different when accompanying chemotherapy or secondary hormonal therapy, the use of corticosteroids was evaluated over 3 study periods: the “overall mCRPC period,” the “secondary hormonal therapy period,” and the “chemotherapy period.” The overall mCRPC period was defined as the entire observation period that spans from index date to the end of eligibility or data availability. The secondary hormonal therapy period was from the first date of treatment with secondary hormonal therapy to the initiation of chemotherapy or the end of eligibility or data availability, whichever occurred first. The chemotherapy period started with the first date of chemotherapy and ended with the start of a secondary hormonal therapy treatment or the end of eligibility or data availability, whichever occurred earlier. If there were more than 1 secondary hormonal therapy or chemotherapy periods for a single patient, the periods were summed to obtain the total secondary hormonal therapy and chemotherapy periods. To describe corticosteroid utilization patterns, the number of patients using corticosteroids (overall and stratified by region), the number of prescriptions per patient, the proportion of days covered, and the length of treatment exposure were reported for each of the periods. The proportion of days covered was calculated as the sum of nonoverlapping days of drug supply in a given period divided by the number of days in that given period. The number of patients receiving prednisone and dexamethasone (stratified by oral and injection administration forms) and the daily prednisone dose were also reported. The daily dose of pharmacy dispensings was obtained by multiplying the medication strength with
310
l
American Health & Drug Benefits
l
the prescribed quantity and then dividing the result by the days of supply.
Statistical analysis Univariate descriptive statistics included the mean (± standard deviation) and the median values for continuous variables and relative frequencies for categorical variables. All analyses were conducted using SAS Version 9.2 (SAS Institute; Cary, NC) or a newer version. Results Population Baseline Characteristics Table 1 (page 311) shows the demographic and clinical baseline characteristics of the study population. The mean age was 73.3 years in population 1 and 68.4 years in population 2. The populations from the 2 databases differed substantially with respect to payer types; although most (73.2%) patients from data set 1 were enrolled in Medicare, almost all (99.5%) patients from data set 2 were covered by commercial insurance. In contrast, the distribution across regions was similar in both databases, with the Midwest and the South accounting for the majority of patients in both databases (38.1% and 32.7% in data set 1 and 31.6% and 42.3% in data set 2, respectively). The mean CCI score was 0.8 in population 1 and 1.0 in population 2. In terms of comorbidities, 19.9% and 24.0% of patients had diabetes in populations 1 and 2, respectively; 35.0% and 53.7%, had hypertension; and 34.4% and 31.8%, had cardiovascular diseases. In both populations, the most common metastatic sites were bone and bone marrow (62.9% in population 1, 63.9% in population 2). The mean observation period lasted 493.3 days in population 1 and 478.0 days in population 2. Evaluation of mCRPC Therapies Table 2 (page 312) shows the distribution of mCRPC therapies in the study populations. The type of first treatment at or after the index date was similar across the study populations, with 35.7% and 31.6% of patients in population 1 and population 2, respectively, undergoing chemotherapy, and 51.7% and 55.9%, respectively, receiving secondary hormonal therapy. Docetaxel was the most common chemotherapy used, accounting for 25.2% of the first treatment and 41.1% of all treatments in population 1, and for 20.3% of the first treatment and 37.5% of all treatments in population 2. Antiandrogens constituted the largest part of secondary hormonal therapies in both populations, with 40.9% of new therapies in data set 1 and 46.0% in data set 2. Bicalutamide was the predominant antiandrogen used, accounting for 37.1% and 40.7% of first treatment and 41.5% and 45.0% of all treatments in population 1 and
www.AHDBonline.com
July/August 2013
l
Vol 6, No 6
Corticosteroid Utilization in Patients with mCRPC
Table 1 Patient Population Baseline Characteristics in the 2 Data Sets Characteristics Age, mean ± SD [median] Age group, N (%)
Original Article
Real-World Corticosteroid Utilization Patterns in Patients with Metastatic Castration-Resistant Prostate Cancer in 2 Large US Administrative Claims Databases Marie-Hélène Lafeuille, MA; Jonathan Gravel, MSc; Amanda Grittner, MA; Patrick Lefebvre, MA; Lorie Ellis, PhD; R. Scott McKenzie, MD Background: Prostate cancer is the most common noncutaneous malignancy in men in the United States. Patients with metastatic castration-resistant prostate cancer (mCRPC) may be treated with secondary hormonal therapy or with chemotherapy, and potentially with concomitant corticosteroids. Corticosteroids can help manage the side effects of chemotherapy and secondary hormonal therapy and ameliorate prostate cancer–related symptoms, although corticosteroids are also associated with adverse effects. With an increasing number of available treatment options for mCRPC, evaluating the real-world concomitant use of corticosteroids in this patient population is important. Objective: To evaluate the utilization patterns of corticosteroids for the treatment of patients with mCRPC based on real-world data from 2 large claim databases. Methods: This retrospective analysis included medical and pharmacy claims from 2 large publicly available healthcare claims databases covering more than 31 million individuals to identify treatment patterns in adult patients with mCRPC. A total of 2593 patients with mCRPC were identified in data set 1 and 626 patients in data set 2 between 2005 and 2011. The appropriate treatment for castration-resistant prostate cancer (CRPC) was defined as chemotherapy, an antiandrogen, an adrenal androgen blocker, or estrogen. The index date was the date of the first CRPC treatment or the first metastasis diagnosis, whichever occurred later. The observation period spanned from the index date to the end of health insurance eligibility. Study end points included population characteristics, the distribution of mCRPC therapies, and corticosteroid utilization patterns. Results: The study population came from the 2 data sets and included 3219 men who were treated for mCRPC. Bone and lymph nodes were the predominant metastatic sites. Bicalutamide was the most common secondary hormonal therapy, and docetaxel was the most common chemotherapy used for these patients. Overall, 73.4% of the patients in data set 1 received concomitant corticosteroids, as did 71.6% of patients in population 2 during the entire period from the index date to the end of eligibility date. In addition, 62.8% and 60.4% of patients, respectively, received concomitant corticosteroids during the secondary hormonal therapy period, and 93.8% and 95.1% of patients, respectively, received concomitant corticosteroids during the chemotherapy period. Similar patterns of corticosteroid use were observed across geographic areas of the United States. Conclusion: This study shows consistently similar utilization patterns of corticosteroids in patients with mCRPC in 2 large national databases. Using real-world data to inform concomitant corticosteroid use in the treatment of patients with mCRPC may assist healthcare providers with treatment selection and with sequencing decision. Future research is warranted to investigate evolving treatment options for patients with mCRPC.
Stakeholder Perspective, page 315
Am Health Drug Benefits. 2013;6(6):307-316 www.AHDBonline.com Disclosures are at end of text
Ms Lafeuille is Senior Economist, Mr Gravel is Economist, Ms Grittner is Economist, and Mr Lefebvre is Vice President, Groupe d’Analyse, Ltée, Montreal, Quebec, Canada; Dr Ellis is Associate Director and Dr McKenzie is Senior Director, Janssen Scientific Affairs, Titusville, NJ.
Vol 6, No 6
l
July/August 2013
www.AHDBonline.com
l
American Health & Drug Benefits
l
307
business
Key Points Corticosteroids can help manage the side effects of chemotherapy or secondary hormonal therapy in patients with metastatic castration-resistant prostate cancer (mCRPC) and ameliorate prostate cancer–related symptoms. ➤ Real-world data on corticosteroid use and treatment patterns are infrequently reported in patients with mCRPC. ➤ Using real-world evidence from medical and pharmacy claims in 2 large claims databases covering >31 million lives, this study identified corticosteroid utilization patterns in patients with mCRPC. ➤ Results show that, in actual practice, the use of concomitant corticosteroid therapy in patients with mCRPC is consistent with the drug’s labeling information across US regions. ➤ In this study, patients with mCRPC were exposed to corticosteroid therapy approximately 33% of the time between 2005 and 2011. ➤ Such real-world data may offer providers positive feedback regarding their treatment selection and sequencing decisions for this patient population. ➤
P
rostate cancer is the most frequently diagnosed noncutaneous malignancy in males and is a leading cause of cancer-related morbidity and mortality among men in the United States.1,2 The American Cancer Society has estimated that approximately 238,600 new cases of prostate cancer will be diagnosed in 2013, and that approximately 29,700 men will die of the disease.3 Many patients will be cured with a primary therapy of surgery, radiation therapy, and potential adjuvant hormonal therapy, but a subset of patients will experience disease progression with metastases spreading to other organs or to tissues in the body, such as to the bladder or bone. For such patients, androgen-deprivation therapy (ADT) is a widely accepted treatment option. Disease progression despite ADT and castrate levels of testosterone is considered castration-resistant prostate cancer (CRPC).4,5 Up to 85% of patients with CRPC are estimated to have metastatic CRPC (mCRPC) and may be considered for secondary hormonal therapy or for chemotherapy.6,7 Historically, treatment for mCRPC was predominantly palliative; however, recent pharmacologic advances have led to increased treatment options showing survival benefits in addition to pain reduction and improvements in quality of life.7,8 An early trial showed that mono
308
l
American Health & Drug Benefits
l
therapy with prednisone may contribute to pain relief and a reduced need for analgesics in a subset of patients.9 In addition, several treatment regimens approved by the US Food and Drug Administration (FDA) for mCRPC involve corticosteroids. Prednisone in combination with mitoxantrone has been associated with pain reduction and with improvement in quality of life in men with advanced, hormone-refractory prostate cancer, but not with improved survival.10 Treatment with docetaxel in combination with prednisone was the first chemotherapy regimen shown to increase survival among patients with mCRPC.10 Men receiving treatment every 3 weeks with docetaxel plus a daily dose of prednisone had a 24% lower probability of death (hazard ratio, 0.76) compared with men treated with mitoxantrone plus prednisone.10 More recent FDA-approved treatment regimens administered with concomitant corticosteroids include abir aterone acetate in combination with prednisone, an oral androgen biosynthesis inhibitor for the treatment of patients with mCRPC, and cabazitaxel in combination with prednisone, a cytotoxic regimen for patients with metastatic hormone-refractory prostate cancer whose disease progresses and who have previously received docetaxel.11,12 Other recently approved treatment regimens without specific corticosteroid requirements include sipuleucel-T, an autologous cellular immunotherapy for asymptomatic or minimally symptomatic patients with mCRPC, and enzalutamide, an androgen receptor inhibitor for patients with progressed mCRPC who have previously received treatment with docetaxel.13,14 Corticosteroids can help manage the side effects of chemotherapy and secondary hormonal therapy and can ameliorate prostate cancer–related symptoms, although they are also associated with adverse effects. An increasing number of treatment options for mCRPC included in the National Comprehensive Cancer Network guidelines involve corticosteroids, but real-world data on corticosteroid use and treatment patterns have been reported infrequently in this patient population.15 The goal of the present study was to address this gap by evaluating the real-world corticosteroid treatment patterns for patients with mCRPC in the United States. A retrospective study design was used to describe the treatment patterns of patients with mCRPC.
Methods Data Source This retrospective study used data from the Truven Health Analytics MarketScan Research Databases (quarter 1 of 2005 through quarter 1 of 2011) and from the Optum Insight Clinformatics Data Mart database (quarter 1 of 2005 through quarter 2 of 2012). The Truven data-
www.AHDBonline.com
July/August 2013
l
Vol 6, No 6
Corticosteroid Utilization in Patients with mCRPC
Figure 1 Study Design
Start of data availability: Q1 of 2005
Start of eligibility
ADT therapy
Index date Initiation of first CRPC therapy or first diagnosis of metastases, whichever was later
Baseline period: 6 mos before index date
End of eligibility
End of data availability: Q1 of 2011a/ Q2 of 2012b
Observation period
Data availability for data set 1 (Truven Health Analytics MarketScan) ended in Q1 of 2011. Data availability for data set 2 (Optum Insight Clinformatics Data Mart) ended in Q2 of 2012. ADT indicates androgen-deprivation therapy; CRPC, castration-resistant prostate cancer; Q, quarter.
a
b
base (data set 1) includes data from 45 large employers, health plans, and government and public organizations covering more than 18 million lives through a commercial insurer or through Medicare. Enrollees in the database include employees, dependents, and retirees in the United States with primary or Medicare supplemental coverage through privately insured fee-for-service, point of service, or capitated health plans. All census regions are represented, although most enrollees are from the South and North Central (ie, Midwest) regions. The Optum database (data set 2) contains claims from more than 13 million privately insured individuals covered by 69 self-insured Fortune 500 companies with locations in all census areas of the United States. Data are available for all company beneficiaries, including employees, spouses, dependents, and retirees. Both databases contain information from integrated, individual-level claims of commercial health insurances, including medical claims for outpatient and inpatient services, pharmacy claims for prescription drugs, and eligibility information on demographics and health plan characteristics. The data were deidentified in compliance with the Health Insurance Portability and Accountability Act of 1996 to preserve patient confidential ity. Institutional Review Board approval and informed consent were not required for this study.
Study Design and Population Selection Adult patients who were treated for mCRPC after ADT, defined as either having at least 1 dispensing of luteinizing hormone-releasing hormone (LHRH) or an orchiectomy, who had at least 1 diagnosis of a malignant neoplasm of the prostate (International Classification of Diseases, Ninth Revision [ICD-9], 185.xx) and at least 1 additional diagnosis of any history of prostate cancer
Vol 6, No 6
l
July/August 2013
(ICD-9, 185.xx or V10.46) within 1 year of at least 1 metastatic diagnosis (ICD-9, 196-199) were included in the study population. Eligible treatments for CRPC included chemotherapy (eg, docetaxel, doxorubicin, estramustine phosphate, etoposide, mitoxantrone, cabazitaxel, paclitaxel, vin blastine, carboplatin, and vinorelbine); antiandrogens (eg, bicalutamide, flutamide, and nilutamide) unless initiated within 90 days of LHRH initiation; adrenal androgen blockers (eg, ketoconazole and aminoglutethimide); and estrogens (eg, diethylstilbestrol, conjugated estrogens, and other estrogens). To ensure that metastases stemmed from prostate cancer, men with another cancer diagnosis (ICD-9, 140.xx-184.xx, 186.xx-209.xx) within 1 year of the first metastatic diagnosis were excluded from the sample. Continuous eligibility for health insurance was required for at least 1 year before the first metastatic diagnosis and at least 6 months before the first CRPC treatment. To obtain a sufficiently long observation period, patients who had an index date less than 3 months before the end of the availability of the data (ie, cutoff date) were excluded from the analysis. Figure 1 illustrates the study design. The index date was defined as either the initiation of the first CRPC therapy or the date of the first diagnosis of metastases, whichever occurred later. The 6 months preceding the index date constituted the baseline period. The observation period spanned from the index date until the end of the individual patient’s health insurance eligibility or until the end of data availability.
Study End Points Population baseline characteristics The demographic and clinical characteristics of the
www.AHDBonline.com
l
American Health & Drug Benefits
l
309
business
study population were evaluated during the baseline period. These characteristics included age, payer type, region, year of index date, clinical history (including score on the Quan-Charlson comorbidity index [CCI]16,17), sites of metastases at the first metastatic diagnosis, and length of observation period.
Evaluation of mCRPC therapies The distribution of mCRPC therapies was evaluated to assess the treatment patterns of patients with mCRPC. The therapies were stratified into secondary hormonal therapies and chemotherapies. Secondary hormonal therapies included antiandrogens, excluding any antiandrogens initiated within 90 days of the initiation of LHRH therapy; adrenal androgen blockers; and estrogen. Relative frequencies for the first treatment of mCRPC and for all treatments during the observation period were calculated. Evaluation of corticosteroid use Because corticosteroid utilization may be different when accompanying chemotherapy or secondary hormonal therapy, the use of corticosteroids was evaluated over 3 study periods: the “overall mCRPC period,” the “secondary hormonal therapy period,” and the “chemotherapy period.” The overall mCRPC period was defined as the entire observation period that spans from index date to the end of eligibility or data availability. The secondary hormonal therapy period was from the first date of treatment with secondary hormonal therapy to the initiation of chemotherapy or the end of eligibility or data availability, whichever occurred first. The chemotherapy period started with the first date of chemotherapy and ended with the start of a secondary hormonal therapy treatment or the end of eligibility or data availability, whichever occurred earlier. If there were more than 1 secondary hormonal therapy or chemotherapy periods for a single patient, the periods were summed to obtain the total secondary hormonal therapy and chemotherapy periods. To describe corticosteroid utilization patterns, the number of patients using corticosteroids (overall and stratified by region), the number of prescriptions per patient, the proportion of days covered, and the length of treatment exposure were reported for each of the periods. The proportion of days covered was calculated as the sum of nonoverlapping days of drug supply in a given period divided by the number of days in that given period. The number of patients receiving prednisone and dexamethasone (stratified by oral and injection administration forms) and the daily prednisone dose were also reported. The daily dose of pharmacy dispensings was obtained by multiplying the medication strength with
310
l
American Health & Drug Benefits
l
the prescribed quantity and then dividing the result by the days of supply.
Statistical analysis Univariate descriptive statistics included the mean (± standard deviation) and the median values for continuous variables and relative frequencies for categorical variables. All analyses were conducted using SAS Version 9.2 (SAS Institute; Cary, NC) or a newer version. Results Population Baseline Characteristics Table 1 (page 311) shows the demographic and clinical baseline characteristics of the study population. The mean age was 73.3 years in population 1 and 68.4 years in population 2. The populations from the 2 databases differed substantially with respect to payer types; although most (73.2%) patients from data set 1 were enrolled in Medicare, almost all (99.5%) patients from data set 2 were covered by commercial insurance. In contrast, the distribution across regions was similar in both databases, with the Midwest and the South accounting for the majority of patients in both databases (38.1% and 32.7% in data set 1 and 31.6% and 42.3% in data set 2, respectively). The mean CCI score was 0.8 in population 1 and 1.0 in population 2. In terms of comorbidities, 19.9% and 24.0% of patients had diabetes in populations 1 and 2, respectively; 35.0% and 53.7%, had hypertension; and 34.4% and 31.8%, had cardiovascular diseases. In both populations, the most common metastatic sites were bone and bone marrow (62.9% in population 1, 63.9% in population 2). The mean observation period lasted 493.3 days in population 1 and 478.0 days in population 2. Evaluation of mCRPC Therapies Table 2 (page 312) shows the distribution of mCRPC therapies in the study populations. The type of first treatment at or after the index date was similar across the study populations, with 35.7% and 31.6% of patients in population 1 and population 2, respectively, undergoing chemotherapy, and 51.7% and 55.9%, respectively, receiving secondary hormonal therapy. Docetaxel was the most common chemotherapy used, accounting for 25.2% of the first treatment and 41.1% of all treatments in population 1, and for 20.3% of the first treatment and 37.5% of all treatments in population 2. Antiandrogens constituted the largest part of secondary hormonal therapies in both populations, with 40.9% of new therapies in data set 1 and 46.0% in data set 2. Bicalutamide was the predominant antiandrogen used, accounting for 37.1% and 40.7% of first treatment and 41.5% and 45.0% of all treatments in population 1 and
www.AHDBonline.com
July/August 2013
l
Vol 6, No 6
Corticosteroid Utilization in Patients with mCRPC
Table 1 Patient Population Baseline Characteristics in the 2 Data Sets Characteristics Age, mean ± SD [median] Age group, N (%)
